Horton

Re: recovering from "found xmin ... from before relfrozenxid ..."

Robert Haas <robertmhaas@gmail.com>

View thread View on postgr.es
From: Robert Haas <robertmhaas@gmail.com>
To: Magnus Hagander <magnus@hagander.net>
Cc: Tom Lane <tgl@sss.pgh.pa.us>, Andres Freund <andres@anarazel.de>, "pgsql-hackers@postgresql.org" <pgsql-hackers@postgresql.org>
Date: 2020-07-14T11:52:35Z
Lists: pgsql-hackers 
On Tue, Jul 14, 2020 at 4:59 AM Magnus Hagander <magnus@hagander.net> wrote:
> The countersable of this is pg_resetwal. The number of people who have broken their database with that tool is not small.

Very true.

> That said, we could have a separate "class" of extensions/tools in the distribution, and encourage packagers to pack them up as separate packages for example. Technically they don't have to be in the same source repository at all of course, but I have a feeling some of them might be a lot easier to maintain if they are. And then the user would just have to install something like "postgresql-14-wizardtools". They'd still be available to everybody, of course, but at least the knives would be in a closed drawer until intentionally picked up.

I don't think that does much to help with the immediate problem here,
because people are being bitten by this problem *now* and a packaging
change like this will take a long time to happen and become standard
out there, but I think it's a good idea.

-- 
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company

Commits

  1. Fix wrong data table horizon computation during backend startup.

  2. Centralize horizon determination for temp tables, fixing bug due to skew.

  3. pg_surgery: Try to stabilize regression tests.

  4. New contrib module, pg_surgery, with heap surgery functions.

  5. Set cutoff xmin more aggressively when vacuuming a temporary table.

  6. snapshot scalability: Don't compute global horizons while building snapshots.

  7. Introduce vacuum errcontext to display additional information.